Multi-Attention-Guided Cascading Network for End-to-End Person Search
The key procedure is to accurately identify pedestrians in complex scenes and effectively embed features from multiple vision cues. However, it is still a limitation to coordinate two tasks in the unified framework, thus leading to high computational overhead and unsatisfactory search performance. F...
Main Authors: | Jianxi Yang, Xiaoyong Wang |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-04-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/13/9/5576 |
Similar Items
-
Attention‐based neural network for end‐to‐end music separation
by: Jing Wang, et al.
Published: (2023-06-01) -
Study on Keyword Search Framework Based on End-to-End Automatic Speech Recognition
by: YANG Run-yan, CHENG Gao-feng, LIU Jian
Published: (2022-01-01) -
Improving Hybrid CTC/Attention Architecture with Time-Restricted Self-Attention CTC for End-to-End Speech Recognition
by: Long Wu, et al.
Published: (2019-10-01) -
Efficient Neural Network for Text Recognition in Natural Scenes Based on End-to-End Multi-Scale Attention Mechanism
by: Huiling Peng, et al.
Published: (2023-03-01) -
An End to End Framework With Adaptive Spatio-Temporal Attention Module for Human Action Recognition
by: Shaocan Liu, et al.
Published: (2020-01-01)